No abstract

The artificial intelligence (AI)-assisted endoscopic detection of early gastric cancer (EGC) has been preliminarily developed. The currently used algorithms still exhibit limitations of large calculation and low-precision expression. The present study aimed to develop an endoscopic automatic detection system in EGC based on a mask region-based convolutional neural network (Mask R-CNN) and to evaluate the performance in controlled trials. For this purpose, a total of 4,471 white light images (WLIs) and 2,662 narrow band images (NBIs) of EGC were obtained for training and testing. In total, 10 of the WLIs (videos) were obtained prospectively to examine the performance of the RCNN system. Furthermore, 400 WLIs were randomly selected for comparison between the Mask R-CNN system and doctors. The evaluation criteria included accuracy, sensitivity, specificity, positive predictive value and negative predictive value. The results revealed that there were no significant differences between the pathological diagnosis with the Mask R-CNN system in the WLI test (χ2 = 0.189, P=0.664; accuracy, 90.25%; sensitivity, 91.06%; specificity, 89.01%) and in the NBI test (χ2 = 0.063, P=0.802; accuracy, 95.12%; sensitivity, 97.59%). Among 10 WLI real-time videos, the speed of the test videos was up to 35 frames/sec, with an accuracy of 90.27%. In a controlled experiment of 400 WLIs, the sensitivity of the Mask R-CNN system was significantly higher than that of experts (χ2 = 7.059, P=0.000; 93.00% VS 80.20%), and the specificity was higher than that of the juniors (χ2 = 9.955, P=0.000, 82.67% VS 71.87%), and the overall accuracy rate was higher than that of the seniors (χ2 = 7.009, P=0.000, 85.25% VS 78.00%). On the whole, the present study demonstrates that the Mask R-CNN system exhibited an excellent performance status for the detection of EGC, particularly for the real-time analysis of WLIs. It may thus be effectively applied to clinical settings.

In the present study, the effects of diabetes on rebleeding following endoscopic treatment were assessed in patients with liver cirrhosis. A retrospective analysis of patients who underwent endoscopic variceal ligation (EVL) or endoscopic injection sclerotherapy (EIS) at the First Affiliated Hospital of Anhui Medical University (Hefei, China) between June 2015 and March 2018 was performed. The patients were divided into the EVL and the EIS groups and each group was subdivided into diabetic and non-diabetic groups. The post-operative rebleeding rate was compared between the EVL and the EIS groups and between the diabetic and non-diabetic patients. The differences in the post-operative rebleeding rate of diabetic patients with hepatogenic and non-hepatogenic diabetes and in patients with different liver function grades were also determined. In the total patient cohort, the rebleeding rate in the EVL subgroup (11.3, 16.5 and 23.5%) was not significantly different compared with that in the EIS subgroup (9.8, 17.4 and 29.3%) at 1, 3 and 6 months following surgery, respectively (P=0.724, 0.868 and 0.339). In the total diabetic group, the rebleeding rate in the EVL subgroup (25.0, 36.1 and 44.4%) was not significantly different compared with that in the EIS subgroup (20.6, 32.4 and 47.1%) at 1, 3 and 6 months following surgery (P=0.660, 0.741 and 0.826, respectively). In the EVL group, the rebleeding rate in the diabetic subgroup (25.0, 36.1 and 44.4%) was higher than that in the non-diabetic subgroup (5.1, 7.6 and 13.9%) at 1, 3, and 6 months following surgery and the differences were significant (P=0.005, <0.001 and <0.001, respectively). In the EIS group, the rebleeding rate in the diabetic subgroup (20.6, 32.4 and 47.1%) was significantly higher than that in the non-diabetic subgroup (3.4, 8.6 and 19.0%) at 1, 3 and 6 months following surgery (P=0.021, 0.004 and 0.004, respectively). Adjustment for age and liver function grade in the EVL and EIS groups was performed using binary logistic regression and the parameter diabetes was indicated to be a risk factor for post-operative rebleeding (P<0.05). No significant difference was noted in the rate of rebleeding between patients with hepatogenic diabetes and non-hepatogenic diabetes at 1, 3 and 6 months following surgery (P=0.634, 0.726 and 0.446, respectively). In the total diabetic group, the rebleeding rate in the Child-Pugh grade A subgroup (14.3, 17.9 and 25.0%) was lower than that in the Child-Pugh grade B/C subgroup (28.6, 45.2 and 59.5%) at 1, 3 and 6 months following surgery, respectively. No significant difference was noted between the two groups at 1 month following surgery (P=0.163). However, the differences were significant at 3 and 6 months following surgery (P=0.018 and 0.005, respectively). The results suggested that diabetes is a risk factor for post-operative rebleeding in patients with cirrhosis. Diabetic patients with poor liver function were more likely to bleed following surgery and the post-operative bleeding rate was not significantly diffe...

Background: The management of large esophageal varices (EVs) remains challenging because of the difficulty of endoscopic variceal ligation and fatal post-endoscopic variceal ligation bleeding ulcers. The current study evaluated the efficacy and safety of balloon-compression endoscopic injection sclerotherapy (bc-EIS) in the treatment of large EVs.Materials and Methods: This retrospective study included 105 patients with cirrhosis exhibiting large EVs (64 in the bc-EIS group and 41 in the EIS group). Primary outcomes included the initial rate of variceal eradication and intraoperative bleeding signs. Secondary outcomes included incidences of rebleeding, mortality, complications, and optimal time of balloon-compression (bc). Results:The initial rate of variceal eradication in the bc-EIS group was significantly higher than that in the EIS group (46.9 vs. 24.4%; P = 0.021). The incidence of intraoperative bleeding, which was represented as oozing and spurting, in the bc-EIS group was markedly lower than that in the EIS group (43.8 vs. 61.0% and 9.4 vs. 39.0%, respectively; P = 0.043). Patients in the bc-EIS group showed a significantly lower incidence of rebleeding (0.0 vs. 17.1%; P = 0.001). However, no significant difference in mortality rate was observed between different groups. Chest pain or discomfort tended to be more common in the EIS group than in the bc-EIS group (58.5 vs. 17.2%; P = 0.001). The cut-off value of 11.5-minutes appeared to have a maximum combined sensitivity and specificity of 80.0% and 58.8%, respectively. The area under the curve was 0.708 (95% confidence interval = 0.576-0.839; P = 0.004).Conclusion: bc-EIS could achieve a higher variceal eradication rate and milder intraoperative bleeding signs in large EVs. Furthermore, 11.5-minutes appeared to be the optimal compression time in bc-EIS.

No abstract